Bayesian Data Assimilation and Uncertainty Quantification
Summer Semester 2022
Notes
- Outline and Bayesian inversion example
- Gaussians
- Probability metrics
- Random dynamical systems I
- Random dynamical systems II
- Smoothing problem
- Filtering problem
- Linear systems I: Kalman filter
- Asymptotics for the (scalar) Kalman filter
- Ill-posedness of (some) inverse problems
- Well-posedness of smoothing and filtering
- MAP estimator and posterior mean
- Linear systems II: Kalman smoother
- Markov Chain Monte Carlo
- Metropolis-Hastings algorithm
- Metropolis-adjusted Langevin algorithm (MALA)
- Independence Sampler and precoditioned Crank-Nicolson scheme
- Example: logistic map
- Variational methods
- Numerical illustration of smoothing algorithms
- Approximate Gaussian filters
- Particle filters
- Convergence of the bootstrap filter
- Variants of the bootstrap filter
Matlab files
- Example #1: linear 2D system
- Example #2: sine map
- Example #3: Lorenz' 63 system
- Example #4: smoothing (deterministic in 1D)
- Example #5: Kalman filter in 2D
- Example #6: linear 1D system
- Example #7: logistic map (smoothing posterior and RWM)
- Example #8: Independence Sampler for the sine map
- Example #9: preconditioned Crank-Nicolson scheme for sine map
- Example #10: 4DVAR for the sine map
- Example #11: particle filter(s) for the sine map
Lecture notes (2019)
Carsten Hartmann, Lorenz Richter. Uncertainty Quantification, 2019 (online).
Projects
- Hamiltonian Monte Carlo (e.g. Neal, MCMC Using Hamiltonian Dynamics, 2011)
- MCMC for functions (e.g. Cotter et al, MCMC Methods for Functions: Modifying Old Algorithms to Make Them Faster, 2013)
- Linear models and regularisation (e.g. Chapter 6.1 in: Sullivan, Introduction to Uncertainty Quantification, 2015)
- Ensemble Kalman Filter (e.g. Katzfuss et al, Understanding the Ensemble Kalman Filter, 2016)
- MAP estimator vs. Bayes estimator (e.g. Bassett & Deride, Maximum a Posteriori Estimators as a Limit of Bayes Estimators, 2019)
- Continuous-time filtering (e.g. Chapter 6.4 in: Law et al, Data Assimilation: A Mathematical Introduction, 2016)
Further reading
- Tim J. Sullivan, Introduction to Uncertainty Quantification, Springer, 2015 (online).
- Sebastian Reich, Colin Cotter, Probabilistic Forecasting and Bayesian Data Assimilation, Cambridge University Press, 2015.
- Kody Law, Andrew Stuart, Kostas Zygalakis, Data Assimilation: A Mathematical Introduction, Springer, 2016 (online).